On this page you can find the publication list of the Machine Learning and Perception Lab.
2016 |
|
Niki Martinel; Claudio Piciarelli; Gian Luca Foresti; Christian Micheloni Mobile Food Recognition with an Extreme Deep Tree Inproceedings International Conference on Distributed Smart Cameras, pp. 56––61, Paris, France, 2016, ISBN: 9781450347860. @inproceedings{Martinel2016c, title = {Mobile Food Recognition with an Extreme Deep Tree}, author = {Niki Martinel and Claudio Piciarelli and Gian Luca Foresti and Christian Micheloni}, doi = {10.1145/2967413.2967428}, isbn = {9781450347860}, year = {2016}, date = {2016-01-01}, booktitle = {International Conference on Distributed Smart Cameras}, pages = {56----61}, address = {Paris, France}, abstract = {Food recognition is an emerging topic in the eld of computer vision. The recent interest of the research community in this area is justi ed by the rise in popularity of food diary applications, where the users take note of their food intake for self-monitoring or to provide useful statistics to dietitians. However, manually annotating food intake can be a tedious task, thus explaining the need of a system that automatically recognizes food, and possibly its amount, from pictures acquired by mobile devices. In this work we propose an approach to food recognition which combines the strengths of di erent state-of-the-art classi ers, namely Convolutional Neural Networks, Extreme Learning Machines and Neural Trees. We show that the proposed architecture can achieve good results even with low computational power, as in the case of mobile devices.}, keywords = {}, pubstate = {published}, tppubtype = {inproceedings} } Food recognition is an emerging topic in the eld of computer vision. The recent interest of the research community in this area is justi ed by the rise in popularity of food diary applications, where the users take note of their food intake for self-monitoring or to provide useful statistics to dietitians. However, manually annotating food intake can be a tedious task, thus explaining the need of a system that automatically recognizes food, and possibly its amount, from pictures acquired by mobile devices. In this work we propose an approach to food recognition which combines the strengths of di erent state-of-the-art classi ers, namely Convolutional Neural Networks, Extreme Learning Machines and Neural Trees. We show that the proposed architecture can achieve good results even with low computational power, as in the case of mobile devices. | |
Niki Martinel; Gian Luca Foresti; Christian Micheloni Distributed and Unsupervised Cost-Driven Person Re-identification Inproceedings International Conference on Pattern Recognition (ICPR), pp. 1225–1230, Cancun, Mexico, 2016. @inproceedings{Martinel2016b, title = {Distributed and Unsupervised Cost-Driven Person Re-identification}, author = {Niki Martinel and Gian Luca Foresti and Christian Micheloni}, url = {https://ieeexplore.ieee.org/document/7899804/}, year = {2016}, date = {2016-01-01}, booktitle = {International Conference on Pattern Recognition (ICPR)}, pages = {1225--1230}, address = {Cancun, Mexico}, abstract = {The problem of re-identify persons across single disjoint camera-pairs has received great attention from the community. Despite this, when the re-identification process has to be carried out on a large camera network a different approach has to be considered. In particular, existing approaches have neglected the importance of the network topology (i.e., the structure of the monitored environment) in such a process. To try filling such a gap, we propose a Distributed and Unsupervised Cost-Driven Person Re-Identification framework (DUPRe) which introduces the following contributions: (i) a camera matching cost to measure the re-identification performance between nodes of the network; (ii) a derivation of the distance vector algorithm which allows to learn the network topology hence to prioritize and limit the cameras inquired for the re-identification. Results on two benchmark datasets show that our solution brings to significant network-wise re-identification improvements.}, keywords = {}, pubstate = {published}, tppubtype = {inproceedings} } The problem of re-identify persons across single disjoint camera-pairs has received great attention from the community. Despite this, when the re-identification process has to be carried out on a large camera network a different approach has to be considered. In particular, existing approaches have neglected the importance of the network topology (i.e., the structure of the monitored environment) in such a process. To try filling such a gap, we propose a Distributed and Unsupervised Cost-Driven Person Re-Identification framework (DUPRe) which introduces the following contributions: (i) a camera matching cost to measure the re-identification performance between nodes of the network; (ii) a derivation of the distance vector algorithm which allows to learn the network topology hence to prioritize and limit the cameras inquired for the re-identification. Results on two benchmark datasets show that our solution brings to significant network-wise re-identification improvements. | |
Niki Martinel; Abir Das; Christian Micheloni; Amit K Roy-Chowdhury Temporal Model Adaptation for Person Re-Identification Inproceedings European Conference on Computer Vision (ECCV), pp. 858–877, Springer, Amsterdam, The Netherlands, 2016. @inproceedings{Martinel2016f, title = {Temporal Model Adaptation for Person Re-Identification}, author = {Niki Martinel and Abir Das and Christian Micheloni and Amit K Roy-Chowdhury}, doi = {10.1007/978-3-319-46493-0_52}, year = {2016}, date = {2016-01-01}, booktitle = {European Conference on Computer Vision (ECCV)}, pages = {858--877}, publisher = {Springer}, address = {Amsterdam, The Netherlands}, abstract = {Person re-identification is an open and challenging problem in computer vision. Majority of the efforts have been spent either to design the best feature representation or to learn the optimal matching metric. Most approaches have neglected the problem of adapting the selected features or the learned model over time. To address such a problem, we propose a temporal model adaptation scheme with human in the loop. We first introduce a similarity-dissimilarity learning method which can be trained in an incremental fashion by means of a stochastic alternating directions methods of multipliers optimization procedure. Then, to achieve temporal adaptation with limited human effort, we exploit a graph-based approach to present the user only the most informative probe-gallery matches that should be used to update the model. Results on three datasets have shown that our approach performs on par or even better than state-of-the-art approaches while reducing the manual pairwise labeling effort by about 80%.}, keywords = {}, pubstate = {published}, tppubtype = {inproceedings} } Person re-identification is an open and challenging problem in computer vision. Majority of the efforts have been spent either to design the best feature representation or to learn the optimal matching metric. Most approaches have neglected the problem of adapting the selected features or the learned model over time. To address such a problem, we propose a temporal model adaptation scheme with human in the loop. We first introduce a similarity-dissimilarity learning method which can be trained in an incremental fashion by means of a stochastic alternating directions methods of multipliers optimization procedure. Then, to achieve temporal adaptation with limited human effort, we exploit a graph-based approach to present the user only the most informative probe-gallery matches that should be used to update the model. Results on three datasets have shown that our approach performs on par or even better than state-of-the-art approaches while reducing the manual pairwise labeling effort by about 80%. | |
Niki Martinel; Christian Micheloni; Gian Luca Foresti A Pool of Multiple Person Re-Identification Experts Journal Article Pattern Recognition Letters, 71 , pp. 23–30, 2016, ISSN: 01678655. @article{Martinel2016g, title = {A Pool of Multiple Person Re-Identification Experts}, author = {Niki Martinel and Christian Micheloni and Gian Luca Foresti}, url = {http://linkinghub.elsevier.com/retrieve/pii/S0167865515004080}, doi = {10.1016/j.patrec.2015.11.022}, issn = {01678655}, year = {2016}, date = {2016-01-01}, journal = {Pattern Recognition Letters}, volume = {71}, pages = {23--30}, publisher = {Elsevier B.V.}, keywords = {}, pubstate = {published}, tppubtype = {article} } | |
2015 |
|
Niki Martinel; Christian Micheloni; Gian Luca Foresti Kernelized Saliency-Based Person Re-Identification Through Multiple Metric Learning Journal Article IEEE Transactions on Image Processing, 24 (12), pp. 5645–5658, 2015, ISSN: 1057-7149. @article{Martinel2015cb, title = {Kernelized Saliency-Based Person Re-Identification Through Multiple Metric Learning}, author = {Niki Martinel and Christian Micheloni and Gian Luca Foresti}, url = {http://ieeexplore.ieee.org/lpdocs/epic03/wrapper.htm?arnumber=7289409}, doi = {10.1109/TIP.2015.2487048}, issn = {1057-7149}, year = {2015}, date = {2015-12-01}, journal = {IEEE Transactions on Image Processing}, volume = {24}, number = {12}, pages = {5645--5658}, abstract = {Person re-identification in a non-overlapping multi-camera scenario is an open and interesting challenge.While the task can hardly be completed by machines, we, as humans, are inherently able to sample those relevant persons' details that allow us to correctly solve the problem in a fraction of a second. Thus, knowing where a human might fixate to recognize a person is of paramount interest for re-identification. Inspired by the human gazing capabilities, we want to identify the salient regions of a person appearance to tackle the problem. Toward this objective, we introduce the following main contributions. A kernelized graph-based approach is used to detect the salient regions of a person appearance, later used as a weighting tool in the feature extraction process. The proposed person representation combines visual features either considering or not the saliency. These are then exploited in a pairwise-based multiple metric learning framework. Finally, the non-Euclidean metrics that have been separately learned for each feature are fused to re-identify a person. The proposed kernelized saliency- based person re-identification through multiple metric learning has been evaluated on four publicly available benchmark data sets to show its superior performance over the state-of-the-art approaches (e.g., it achieves a rank 1 correct recognition rate of 42.41% on the VIPeR dataset).}, keywords = {}, pubstate = {published}, tppubtype = {article} } Person re-identification in a non-overlapping multi-camera scenario is an open and interesting challenge.While the task can hardly be completed by machines, we, as humans, are inherently able to sample those relevant persons' details that allow us to correctly solve the problem in a fraction of a second. Thus, knowing where a human might fixate to recognize a person is of paramount interest for re-identification. Inspired by the human gazing capabilities, we want to identify the salient regions of a person appearance to tackle the problem. Toward this objective, we introduce the following main contributions. A kernelized graph-based approach is used to detect the salient regions of a person appearance, later used as a weighting tool in the feature extraction process. The proposed person representation combines visual features either considering or not the saliency. These are then exploited in a pairwise-based multiple metric learning framework. Finally, the non-Euclidean metrics that have been separately learned for each feature are fused to re-identify a person. The proposed kernelized saliency- based person re-identification through multiple metric learning has been evaluated on four publicly available benchmark data sets to show its superior performance over the state-of-the-art approaches (e.g., it achieves a rank 1 correct recognition rate of 42.41% on the VIPeR dataset). | |
Niki Martinel; Abir Das; Christian Micheloni; Amit K Roy-Chowdhury Re-Identification in the Function Space of Feature Warps Journal Article IEEE Transactions on Pattern Analysis and Machine Intelligence (TPAMI), 37 (8), pp. 1656–1669, 2015, ISSN: 0162-8828. @article{Martinel2015a, title = {Re-Identification in the Function Space of Feature Warps}, author = {Niki Martinel and Abir Das and Christian Micheloni and Amit K Roy-Chowdhury}, url = {http://ieeexplore.ieee.org/lpdocs/epic03/wrapper.htm?arnumber=6975169}, doi = {10.1109/TPAMI.2014.2377748}, issn = {0162-8828}, year = {2015}, date = {2015-08-01}, journal = {IEEE Transactions on Pattern Analysis and Machine Intelligence (TPAMI)}, volume = {37}, number = {8}, pages = {1656--1669}, keywords = {}, pubstate = {published}, tppubtype = {article} } | |
Niki Martinel; Christian Micheloni; Gian Luca Foresti The Evolution of Neural Learning Systems: A Novel Architecture Combining the Strengths of NTs, CNNs, and ELMs Journal Article IEEE Systems, Man, and Cybernetics Magazine, 1 (3), pp. 17–26, 2015, ISSN: 2333-942X. @article{Martinel2015b, title = {The Evolution of Neural Learning Systems: A Novel Architecture Combining the Strengths of NTs, CNNs, and ELMs}, author = {Niki Martinel and Christian Micheloni and Gian Luca Foresti}, url = {http://ieeexplore.ieee.org/lpdocs/epic03/wrapper.htm?arnumber=7426555}, doi = {10.1109/MSMC.2015.2461151}, issn = {2333-942X}, year = {2015}, date = {2015-07-01}, journal = {IEEE Systems, Man, and Cybernetics Magazine}, volume = {1}, number = {3}, pages = {17--26}, keywords = {}, pubstate = {published}, tppubtype = {article} } | |
Niki Martinel; Christian Micheloni Classification of Local Eigen-Dissimilarities for Person Re-Identification Journal Article IEEE Signal Processing Letters, 22 (4), pp. 455–459, 2015, ISSN: 1070-9908. @article{Martinel2014, title = {Classification of Local Eigen-Dissimilarities for Person Re-Identification}, author = {Niki Martinel and Christian Micheloni}, url = {http://ieeexplore.ieee.org/lpdocs/epic03/wrapper.htm?arnumber=6919267 http://ieeexplore.ieee.org/document/6919267/}, doi = {10.1109/LSP.2014.2362573}, issn = {1070-9908}, year = {2015}, date = {2015-04-01}, journal = {IEEE Signal Processing Letters}, volume = {22}, number = {4}, pages = {455--459}, keywords = {}, pubstate = {published}, tppubtype = {article} } | |
N Martinel; C Piciarelli; C Micheloni; G L Foresti On filter banks of texture features for mobile food classification Inproceedings ACM International Conference Proceeding Series, 2015, ISBN: 9781450336819. @inproceedings{Martinel2015g, title = {On filter banks of texture features for mobile food classification}, author = {N Martinel and C Piciarelli and C Micheloni and G L Foresti}, url = {https://dl.acm.org/doi/abs/10.1145/2789116.2789132}, doi = {10.1145/2789116.2789132}, isbn = {9781450336819}, year = {2015}, date = {2015-01-01}, booktitle = {ACM International Conference Proceeding Series}, volume = {08-11-Sep-}, abstract = {textcopyright 2015 ACM. Nowadays obesity has become one of the most common diseases in many countries. To face it, obese people should constantly monitor their daily meals both for self-limitation and to provide useful statistics for their dietitians. This has led to the recent rise in popularity of food diary applications on mobile devices, where the users can manually annotate their food intake. To overcome the tediousness of such a process, several works on automatic image food recognition have been proposed, typically based on texture features extraction and classification. In this work, we analyze different texture filter banks to evaluate their performances and propose a method to automatically aggregate the best features for food classification purposes. Particular emphasis is put in the computational burden of the system to match the limited capabilities of mobile devices.}, keywords = {}, pubstate = {published}, tppubtype = {inproceedings} } textcopyright 2015 ACM. Nowadays obesity has become one of the most common diseases in many countries. To face it, obese people should constantly monitor their daily meals both for self-limitation and to provide useful statistics for their dietitians. This has led to the recent rise in popularity of food diary applications on mobile devices, where the users can manually annotate their food intake. To overcome the tediousness of such a process, several works on automatic image food recognition have been proposed, typically based on texture features extraction and classification. In this work, we analyze different texture filter banks to evaluate their performances and propose a method to automatically aggregate the best features for food classification purposes. Particular emphasis is put in the computational burden of the system to match the limited capabilities of mobile devices. | |
J Garcia; N Martinel; C Micheloni; A Gardel Person re-identification ranking optimisation by discriminant context information analysis Inproceedings Proceedings of the International Conference on Computer Vision (ICCV), 2015, ISSN: 15505499. @inproceedings{Garcia2015a, title = {Person re-identification ranking optimisation by discriminant context information analysis}, author = {J Garcia and N Martinel and C Micheloni and A Gardel}, doi = {10.1109/ICCV.2015.154}, issn = {15505499}, year = {2015}, date = {2015-01-01}, booktitle = {Proceedings of the International Conference on Computer Vision (ICCV)}, volume = {2015 Inter}, abstract = {textcopyright 2015 IEEE. Person re-identification is an open and challenging problem in computer vision. Existing re-identification approaches focus on optimal methods for features matching (e.g., metric learning approaches) or study the inter-camera transformations of such features. These methods hardly ever pay attention to the problem of visual ambiguities shared between the first ranks. In this paper, we focus on such a problem and introduce an unsupervised ranking optimization approach based on discriminant context information analysis. The proposed approach refines a given initial ranking by removing the visual ambiguities common to first ranks. This is achieved by analyzing their content and context information. Extensive experiments on three publicly available benchmark datasets and different baseline methods have been conducted. Results demonstrate a remarkable improvement in the first positions of the ranking. Regardless of the selected dataset, state-of-the-art methods are strongly outperformed by our method.}, keywords = {}, pubstate = {published}, tppubtype = {inproceedings} } textcopyright 2015 IEEE. Person re-identification is an open and challenging problem in computer vision. Existing re-identification approaches focus on optimal methods for features matching (e.g., metric learning approaches) or study the inter-camera transformations of such features. These methods hardly ever pay attention to the problem of visual ambiguities shared between the first ranks. In this paper, we focus on such a problem and introduce an unsupervised ranking optimization approach based on discriminant context information analysis. The proposed approach refines a given initial ranking by removing the visual ambiguities common to first ranks. This is achieved by analyzing their content and context information. Extensive experiments on three publicly available benchmark datasets and different baseline methods have been conducted. Results demonstrate a remarkable improvement in the first positions of the ranking. Regardless of the selected dataset, state-of-the-art methods are strongly outperformed by our method. | |
Niki Martinel; Claudio Piciarelli; Christian Micheloni; Gian Luca Foresti A Structured Committee for Food Recognition Inproceedings Proceedings of the IEEE International Conference on Computer Vision (ICCV) Workshops, pp. 92–100, Santiago, Chile, 2015. @inproceedings{Martinel2015c, title = {A Structured Committee for Food Recognition}, author = {Niki Martinel and Claudio Piciarelli and Christian Micheloni and Gian Luca Foresti}, url = {http://openaccess.thecvf.com/content_iccv_2015_workshops/w12/html/Martinel_A_Structured_Committee_ICCV_2015_paper.html}, year = {2015}, date = {2015-01-01}, booktitle = {Proceedings of the IEEE International Conference on Computer Vision (ICCV) Workshops}, pages = {92--100}, address = {Santiago, Chile}, keywords = {}, pubstate = {published}, tppubtype = {inproceedings} } | |
Niki Martinel; Claudio Piciarelli; Christian Micheloni; Gian Luca Foresti On Filter Banks of Texture Features for Mobile Food Classification Inproceedings International Conference on Distributed Smart Cameras, pp. 11–16, Seville, Spain, 2015, ISBN: 9781450336819. @inproceedings{Martinel2015e, title = {On Filter Banks of Texture Features for Mobile Food Classification}, author = {Niki Martinel and Claudio Piciarelli and Christian Micheloni and Gian Luca Foresti}, url = {https://dl.acm.org/doi/10.1145/2789116.2789132}, isbn = {9781450336819}, year = {2015}, date = {2015-01-01}, booktitle = {International Conference on Distributed Smart Cameras}, pages = {11--16}, address = {Seville, Spain}, keywords = {}, pubstate = {published}, tppubtype = {inproceedings} } | |